Recent evidence suggests that tumors grow in hierarchies driven by distinct cells with the ability to self renew and differentiate, called cancer stem cells (CSCs). The ability to study CSCs in solid tumors has been hampered by a lack of simple mouse models that mimic human disease. We recently discovered mouse pancreatic CSCs that, when introduced into histocompatible recipients with normal immune systems, generate pancreatic tumors that mimic human pancreatic ductal adenocarcinoma (PDA) in terms of histologic appearance and pattern of disease progression. The objective of this project is to study the immune response to murine pancreatic CSCs and identify molecules that distinguish these cells from non-CSC tumor cells. We recently discovered that in the LSL-KrasG12D/+; LSL-p53 R172H/+; Pdx-1-Cre mouse model of pancreatic cancer, cells isolated from liver metastases grow in vitro and display CSC markers. Moreover, as few as 500 cells injected into immunocompetent, histocompatible mice can generate differentiated tumors that are histologically indistinguishable from human pancreatic adenocarcinoma. Primary tumors as well as sites of metastases are rapidly infiltrated with host immune cells. We will 1) generate clonogenic murine pancreatic CSC lines from primary and metastatic tumor cells and characterize these lines with respect to their morphology, chromosomal stability, and cancer stem cell properties; 2) characterize the cellular immune response to CSCs and non-CSCs in tumor-bearing and tumor-naive mice; 3) use gene profiling to identify genes that are upregulated in CSCs versus non-CSC tumor cells; 4) evaluate the ability of short hairpin RNA (shRNA) agents directed at molecules expressed by CSCs to block the growth of these cells in vitro and their generation of tumors in vivo; and 5) evaluate the role of CXCR4 in CSC growth and metastasis and assess the effects of host cells, including immune cells, on CXCR4 expression. The findings from this project should elucidate the biology of pancreatic cancer and lead to the identification of new therapeutic targets for this devastating disease. RELEVANCE: This model, which closely mimics human pancreatic cancer, provides a novel means to study pancreatic CSCs. Since our CSCs have the same genetic mutations (Kras, p53) commonly found in pancreatic cancer and express the same molecular markers as human pancreatic CSCs, knowledge gained about the biology of these cells should prove useful in the design of more effective therapies for human pancreatic cancer.